Microwave irradiation and catalysis in organophosphorus reactions

Introduction

The spread of the microwave (MW) technique has had a great impact also on organophosphorus (OP) chemistry. Guenin was who summarized the initial results [1]. The senior author of this paper and his co-workers made efforts to make use of the MW tool in the synthesis of P-heterocycles and other phosphinic-, phosphonic- and phosphine oxide derivatives [2], [3], [4], [5], [6], [7], [8], [9]. The stress was placed on green chemical aspects [3], [4]. Beside elaborating efficient syntheses characterized by short reaction times and good yields, it was also possible to promote otherwise reluctant transformations, such as the inverse-Wittig type reaction of 2,4,6-trialkylphenylphosphine oxides with dialkyl acetylenedicarboxylates to afford β-oxophosphoranes [10], [11], [12]. Another challenge was for us to evaluate the scope and limitations of the use of the MW technique in organic syntheses, and to model the effect of MWs [13], [14], [15].

In this article, we wish to summarize our results on the advantageous use of catalysts in MW-assisted reactions, and on the beneficial effect of MWs in a catalytic transformation. Furthermore, we show cases, when catalysts may be simplified, or even omitted under MW irradiation.

MW-assisted direct esterification of cyclic phosphinic acids – the beneficial effect of an ionic liquid additive

We found that cyclic phosphinic acids, such as 1-hydroxy-3-phospholene oxides (1), 1-hydroxyphospholane oxides (2) and a 1-hydroxy-1,2,3,4,5,6-hexahydrophosphinine oxide (3) underwent direct esterification with simple alcohols used in a 15-fold excess at 220–235°C under MW irradiation to give the corresponding phosphinates (4–6) (Scheme 1) [16], [17], [18].

Scheme 1:

MW-assisted esterification of 5- and 6-ring phosphinic acids (1–3).

It was experienced that the esterifications with longer carbon atom chain (and hence less volatile) alcohols were more efficient giving the corresponding phosphinates (4–6) in yields of 60–95%. The reaction times were mostly around 1–5 h. We also observed that the MW-promoted thioesterifications and amidations led only to incomplete conversions [19], [20].

These days, ionic liquids are found applications as “green” solvents [21], but they might be proven even more useful as catalysts or additives [22]. We experienced that in the presence of 10% of [bmim][PF₆], the MW-assisted esterification of 1-hydroxy-3-methyl-3-phospholene oxide (1A) could be performed at a lower temperature (180°C), in a shorter reaction time (0.5 h), and in a high conversion (90%) allowing a good yield (83%), as compared to the MW-assisted variation lacking an IL additive, in which case a conversion of 60% could be attained after an irradiation at 200°C for 2 h (Scheme 2) [23], [24]. For comparison purposes, results of the inefficient thermal experiments are also shown in Scheme 2.

Scheme 2:

Thermal and MW-assisted esterification of 1-hydroxy-3-methyl-3-phospholene oxide (1A) with butanol in the absence or presence of an ionic liquid additive.

The ionic liquid catalysis was then utilized in the esterification of phosphinic acid 1A with other alcohols like ethanol, propanol, pentanol and isooctanol (Table 1). The effect of [bmim][PF₆] was indeed significant in promoting higher conversions and yields at lower temperatures and after shorter reaction times. It is an especially valuable result that the esterifications with volatile alcohols like EtOH and PrOH became much more efficient [23], [24].

Table 1:

MW-assisted esterification of 1-hydroxy-3-methyl-3-phospholene oxide (1A) with different alcohols in the absence or presence of [bmim][PF₆].

R	[bmim][PF6] (%)	T (°C)	p (bar)	t (h)	Conversion (%)	Yield (%)
Et (b)	–	160	17	4	38	30
Et (b)	10	160	17	3	86	60
Pr (c)	–	180	15.5	4	40	30
Pr (c)	10	180	15.5	3	98	68
Pent (d)	–	220	9	2.5	100	82
Pent (d)	10	180	5	0.5	100	94
i Oct (e)	–	220	3	1	100	76
i Oct (e)	10	180	2	0.33	100	84

And what is more, the [bmim][[PF₆]-promoted protocol could be extended to the esterification of dimethyl-hydroxyphospholene oxide 1B, 1-hydroxyphospholane oxides 2A and 2B, and a hydroxy-1,2,3,4,5,6-hexahydrophosphinine oxide (3) using pentanol as the alcohol (Table 2). The significant effect of the IL additive could again be observed [23], [24].

Table 2:

Extension of the MW-assisted IL-catalyzed direct esterification to other cyclic phosphinic acids (1B, 2A, 2B and 3).

Model reaction	[bmim][PF6] (%)	T (°C)	p (bar)	t (h)	Yield (%)
	–	235	11	3	67
	10	200	6	1	72
	–	235	11	3	79
	10	220	9	1	89
	–	235	11	5	60
	10	220	9	2	84
	–	220	9	4	31
	10	200	9	2	42

MW-assisted alkylating esterification of cyclic phosphinic acids – a phase transfer catalytic support

The synergism of MW irradiation and phase transfer catalysis was observed in the alkylating esterification of cyclic phosphinic acids, such as 1-hydroxy-3-phospholene oxides (1), 1-hydroxyphospholane oxides (2) and 1-hydroxy-1,2,3,4,5,6-hexahydrophosphinine oxide (3) in the presence of K₂CO₃ under solvent-free conditions using n-alkyl halides (Scheme 3). The phosphinates (4–6) were obtained in yields of 73–95% [16], [25], [26]. In the case when an alkyl halide with increased reactivity, such as benzyl bromide was used, MW was beneficial alone, and there was no need for a catalyst.

Scheme 3:

Alkylating esterification of cyclic phosphinic acids (1–3).

Catalytic Michaelis–Arbuzov reactions – MW assistance

The Arbuzov reaction is a suitable approach for the preparation of arylphosphonates [27]. Special catalytic protocols are necessary to overcome the low reactivity of the aryl halides. The Ni-catalyzed reaction of aryl bromides with triethyl phosphite was further developed, and performed under MW irradiation. The phosphonates (7) could be prepared in yields of 67–90% (Table 3) [28].

Table 3:

MW-assisted Arbuzov reaction of aryl halides and triethyl phosphite.

R1	R2	R3	R4	Yield (%)
H	H	H	H	90
CH3	H	H	H	86
H	H	CH3	H	71
H	H	Cyclohexyl	H	86
H	H	OCF3	H	74
H	H	CF3	H	72
H	H	OCH3	H	69
H	H	C(O)CH3	H	73
F	H	H	H	67
H	F	H	H	82
H	H	F	H	73
F	H	H	F	78

Alkylation of active methylene-containing compounds – replacement of the phase transfer catalyst by MW irradiation

A typical case, where the catalyst may be substituted by MW irradiation is the alkylation of CH acidic compounds, such as tetraethyl methylenebisphosphonate (8), diethyl cyanomethylphosphonate (9), and diethyl ethoxycarbonylmethylphosphonate (10) (Table 4). In a solid–liquid phase accomplishment, potassium- or cesium carbonate was used as the base, and under MW conditions, there was no need for phase transfer catalyst. The monoalkylated products (11–13) were obtained in yields of 64–80% [29], [30], [31]. A dialkylating protocol was also elaborated by us [32].

Table 4:

Catalyst-free alkylation of P=O-functionalized active methylene containing compounds (8–10) under MW irradiation in solid–liquid phase.

R1	Starting material	R	M	T (°C)	Product	Yield (%)
P(O)(OEt)2	8	Et	Cs	140	11	~80
CN	9	Pr	K	100	12	64
CO2Et	10	Bu	Cs	120	13	70

The Kabachnik–Fields reaction – substitution of the catalyst by MW irradiation

α-Aminophosphonates (14, Y=RO), and α-aminophosphine oxides (14, Y=Ph) may be synthesized by the solvent- and catalyst-free MW-assisted Kabachnik–Fields (phospha-Mannich) condensation of primary amines, oxo compounds and >P(O)H reagents, such as dialkyl phosphites and diphenylphosphine oxide. Earlier preparations applied special catalysts [33], [34], [35], [36], [37], which mean cost and environmental burden. It was found that under MW conditions, there is no need for any catalyst (Scheme 4) [38].

Scheme 4:

Catalyst-free Kabachnik–Fields reaction under MW irradiation.

The use of heterocyclic amines, such as pyrrolidine, piperidine derivatives, morpholine and piperazine derivatives, or heterocyclic >P(O)H species (e.g. 1,3,2-dioxaphosphorine oxide) led to N-heterocyclic [39] and P-heterocyclic [40] α-aminophosphonates. 3-Amino-6-methyl-2H-pyran-2-ones could also serve as amino components in Kabachnik–Fields reaction with formaldehyde and dialkyl phosphites or diphenylphosphine oxide [41].

Primary amines are able to participate in bis(Kabachnik–Fields) condensations [42]. In such cases, alkyl or arylamines were reacted with two equivalents of the formaldehyde and the >P(O)H species to give the bis(Z¹Z²P(O)CH₂)amines (15) (Scheme 5) [43], [44], [45]. Most of the reactions could be carried out without the use of solvent.

Scheme 5:

MW-assisted bis(Kabachnik–Fields) reaction.

As further developments, the bisphosphinoyl derivatives (15, Z¹=Z²=Ph) were deoxygenated to bis(phosphines) that were converted to cyclic platinum complexes [44], [45], [46]. α-, β- And γ-amino acids (or their esters) were also utilized as starting materials in the double Kabachnik–Fields condensation to provide the corresponding bis(phosphono- or phosphinoyl) products [47], [48].

As special bis(Kabachnik–Fields) reactions, paraphenylene diamine was reacted with two equivalents of substituted benzaldehydes and diethyl phosphite, or terephthalaldehyde was reacted with two equivalents of an arylamine and a P-reagent [49].

Deoxygenation of phosphine oxides – MW as a substitute for catalyst

As was mentioned in the part discussing the phospha-Mannich reaction [44], [45], [46], the reduction of phosphine oxides to phosphines is an important transformation that is usually realized by the use of silanes [50]. We found that under MW and solvent-free conditions, there is no need to use any catalyst. Moreover, user-friendly silanes, such as tetramethyl-disiloxane (TMDS) or polymethoxysiloxane (PMHS) could also be applied [51], [52], [53], [54]. Selected results are shown on the reduction of 1-phenyl-3-methyl-3-phospholene oxide 16 (Table 5). Deoxygenation is a usual protocol to make available racemic or optically active P-ligands from the corresponding P-oxide precursors [50], [55].

Table 5:

Catalyst-free deoxygenation of 1-phenyl-3-methyl-3-phospholene 1-oxide by different silanes.

Silane	Equivalents	Mode of heating	T (°C)	t (h)	Conversion (%)	Yield (%)
PhSiH3	3	Δ	80	2	100	95
PhSiH3	3	MW	80	1	100	91
TMDS	4	Δ	110	5	100	92
TMDS	4	MW	110	3	100	92
PMHS	2	Δ	110	4	100	91
PMHS	2	MW	110	2	100	92

The Hirao reaction – A “P-ligand-free” accomplishment under MW irradiation

The Hirao reaction involving the P–C coupling of a vinyl halide/aryl halide/hetaryl halide and a >P(O)H reagent has become an important tool in the synthesis of phosphonates, phosphinates and phosphine oxides [56], [57]. We found that in the coupling reaction of dialkyl phosphites with bromobenzene, there was no need for the expensive Pd(Ph₃P)₄ catalyst, or for the combined addition of Pd(OAc)₂ and a P-ligand, as Pd(OAc)₂ also catalyses the Hirao reaction in the in the presence of the >P(O)H reagent. A MW-assisted solvent-free accomplishment was the best choice [58], [59]. Arylphosphonates 18 were obtained in 69–93% yields (Scheme 6).

Scheme 6:

“P-ligand-free” MW-assisted coupling reaction of dialkyl phosphites and aryl bromides.

The similar reaction of alkyl phenyl-H-phosphinates with bromobenzene led to alkyl diphenylphosphinates (19) (Scheme 7/(1)), while the reaction of secondary phosphine oxides with bromoarenes afforded phosphine oxides (20) (Scheme 7/(2)).

Scheme 7:

“P-ligand-free” MW-assisted coupling reaction of alkyl phenyl-H-phosphinates and secondary phosphine oxides with bromoarenes.

Furthermore, we found that the MW-assisted “P-ligand-free” protocol worked also with NiCl₂ as the catalyst in the synthesis of arylphosphonates (18), alkyl diphenylphosphinates (19) and diaryl-phenylphosphine oxides (20, Z=Ar). In these variations, acetonitrile had to be used due to the heterogeneity of the reaction mixtures [60].

The P–C coupling of halobenzoic acids and diarylphosphine oxides could be performed in the absence of any catalyst, in the presence of K₂CO₃ in water under MW conditions (Scheme 8) [61].

Scheme 8:

Catalyst-free coupling reaction of diarylphosphine oxides and halobenzoic acids in water.

As a matter of fact, in the “P-ligand-free” cases outlined above, the trivalent tautomer form of the >P(O)H reagent acts as the ligand [62].

The classical catalytic cycle for the Hirao reaction involves oxidative addition of the aryl halide to the Pd(0) complex (22) to give Pd(II) adduct 23, the change of ligands leading to key intermediate 24, and finally the reductive elimination affording the product (ArP(O)Y₂) (Fig. 1).

Fig. 1:

Classical catalytic cycle for the Pd-catalyzed Hirao reaction.

The 23→24 transformation may be refined assuming the complexation of the trivalent form of the Y₂P(O)H reagent, and that that this step is followed by deprotonation (Scheme 9).

Scheme 9:

Refinement of the ligand change step in the general catalytic cycle.

As regards the interaction of Pd(OAc)₂ and the >P(O)H species, Buono et al. claimed the formation of dimer type complexes (26) shown in Scheme 10 [63], [64].

Scheme 10:

Assumed formation of bis(palladium complexes) 26 from Pd(OAc)₂ and secondary phosphine oxides.

It is noteworthy that the formation of an analogous dimer (27) was observed in the reaction of PdCl₂(MeCN)₂ with tert-butyl-phenylphosphine oxide (Scheme 11).

Scheme 11:

Formation of bis(palladium complex) 27 from PdCl₂(MeCN)₂ and tert-butyl-phenylphosphine oxide.

Theoretical calculations suggested that the reaction of Pd(OAc)₂ with diphenylphosphine oxide or dimethyl phosphite cannot give the corresponding dimer type complexes 28, as their formation from the corresponding monomers is highly unfavorable marked by the enthalpy change of +82.3 kJ mol⁻¹ and +20.7 kJ mol⁻¹, respectively (Scheme 12) [62].

Scheme 12:

Assumed formation of dimer type Pd complexes 28 from Pd(OAc)₂ and diphenylphosphine oxide or dimethyl phosphite.

However, the analogous formation of the chloro-containing dimer type complexes 29 from the corresponding monomers was justified, as their formation is exothermic as characterized by enthalpy changes of –118.9 kJ mol⁻¹ and –153.8 kJ mol⁻¹, respectively (Scheme 13) [62].

Scheme 13:

Formation of bis(palladium complexes) 29 from PdCl₂ and diphenylphosphine oxide or dimethyl phosphite.

Our idea was that the interaction of the Y₂P(O)H species and Pd(OAc)₂ results in eventually Pd(PY₂OH)₂ complex (31). In the first part of this series of reactions, the Pd²⁺ ion of Pd(OAc)₂ is reduced to Pd⁰ by reaction with Y₂P(O)H, then the metal is complexed by two Y₂POH units (Scheme 14). As a newer development, intermediate 30 could be pointed out by theoretical calculations (Fig. 2).

Scheme 14:

A realistic proposal for the Pd complex formed from Pd(OAc)₂ and Y₂P(O)H.

Fig. 2:

Intermediate 30 (Y = MeO) substantiated in the reaction of Pd(OAc)₂ with (MeO)₂P(O)H using the G09 program [65] at the B3LYP level of theory with basis set 6-31G(d,p) for CHOP [66], and SDD/MWB28 for Pd [67]. [Selected geometries. Bond distances in Å: P–OMe1: 1.61, P–OMe2: 1.60, P–OH: 1.58, P–Pd: 2.32, Atomic distances in Å: P· · ·O: 2.75, Pd· · ·O1: 1.98, Pd· · ·O2: 2.04, Pd· · ·O3: 2.42, Bond angles in deg: Pd–P–OC1: 119.1, Pd–P–OC2: 120.4, Pd–P–OH: 100.7, P–O–C1: 128.0, P–O–C2: 129.5, P–O–H: 115.0, Dihedral angles in deg: Pd–P–O–C1: 130.1, Pd–P–O–C2: 18.3, Pd–P–O–H: 4.9].

The formation of complex 31 was in full agreement with our experimental data. The best experiment was, when 1.3 equivalents of the >P(O)H species were applied together with 10% of the Pd(OAc)₂ catalyst. (0.1 Equivalents served as the reducing agent, 0.2 equivalents provided the P-ligand, and one equivalent was consumed as the reagent [62]).

High level quantum chemical calculations were performed on the “P-ligand-free” Hirao reaction of bromobenzene and diethyl phosphite or diphenylphosphine oxide (Scheme 15).

Scheme 15:

The model reaction calculated by the B3LYP/6-31G(d,p) [66] method applying the PCM solvent model [62].

In the first stage of the transformation, catalyst complex 30 undergoes oxidative addition of PhBr in three elemental steps to afford species 32 (see top of Fig. 3). Then Pd-complex 32 loses the bromide anion to give species 33 that by reaction with Y₂POH (1′) affords complex 34. Deprotonation and reductive rearrangement via intermediate 35 and the corresponding TS, respectively, led to complex 36, whose decomposition furnished the P–C coupled product 31, and regenerated catalyst 30 (Fig. 2) [62].

Fig. 3:

Catalytic cycle for the Hirao reaction obtained. Computations were carried out by using G09 program [65] at the B3LYP level of theory with the /6-31G(d,p) basis set for nuclei CHNOPBr [66], and MWB28 for Pd [67], applying the PCM solvent model [68].

The “P-ligand-free” approaches are environmentally-friendly as save costs and environmental burdens. Till data, the P-ligand-free variation may be the most attractive protocol for the Hirao reaction. The “green” accomplishments have been reviewed by us [56], [57].

The Pudovik reaction – no need for MW assistance, as triethylamine-catalysis is enough

α-Aryl-α-hydroxyphosponates and α-aryl-α-hydroxyphosphine oxides (37) were synthesized in a catalytic and solvent-free MW-assisted reaction involving the addition of >P(O)H species to aryl aldehydes (Scheme 16) [69].

Scheme 16:

MW-assisted synthesis of α-hydroxyphosphonates and α-hydroxyphosphine oxides.

Dialkyl phosphites were also added to the carbonyl function of α-ketophosphonates in the presence of diethylamine and in the absence of any solvent under MW conditions to result in the formation of dronate analogs, α-hydroxybisphosphonates [70], [71].

However, MW assistance is not definitely necessary in the case of Pudovik reactions. We found that in the presence of trimethylamine as the catalyst dialkyl phosphites add on the carbonyl group of benzaldehyde derivatives already in acetone at the boiling point. Moreover, the α-aryl-α-hydroxyphosphonates precipitate on cooling from the mixture, hence this method represents the “greenest” protocol for the preparation of α-aryl-α-hydroxyphosphonates [72].

To our surprise, the α-hydroxyphosphonates prepared could be easily transformed to the corresponding α-aminophosphonates by reaction with primary amines [73]. The substitution reaction was promoted by the neighboring group effect of the adjacent P=O function.

Summary

In summary, typical cases were overviewed for the MW versus catalytic accomplishments. There are instances (esterification of phosphinic acids), when MW assistance is further promoted by a suitable catalyst, or the opposite situation, when a catalytic transformation (e.g. a hindered Arbuzov reaction) is enhanced by MW irradiation. An additional possibility is, when a catalyst may be simply omitted under MW conditions as represented by the phospha-Mannich reaction, special C-alkylations, or the reduction of phosphine oxides. Last but not least, it was found that under MW irradiation, the excess of the >P(O)H reagent may serve as the P-ligand, hence there is no need to add the usual P(III) species.